Overview
这本书是对《Science Research Writing For Non-Native Speakers of English》的总结梳理。这本书分为多个章节，每个章节对应一个论文写作里的部分。在每个章节内部，又有如下几个部分：
- 介绍一下这个章节在整体文章中的作用
- 展示一篇例文
- 介绍一些基础的英语语法和词汇
- 对例文结构进行分析（分解成小组件），也就是他说的 model
- 对 model 进行一些心法层面的阐述，也就是他说的 key
- 按照 model 和 key 介绍一些固定词汇、句式和搭配，也就是他说的 vocabulary
- 最后让你自己写一个文章
很多东西都是实践性质的，并非专业知识本身。所以如果总结的话，其实可以按照 model 的形式来组织 vocabulary ，其他的东西（比如例文、基础语法、动手实践）都可以省略，结构清晰。
因为时间问题，所以我会逐步更新这篇博文。
Architecture
这篇文章将一片论文分成了 5 个部分：
- Abstract 用于提供一个对文章的总结
- Introduction 初步介绍问题，帮助原本陌生的读者对问题的认识更加具象和熟悉
- Methodology 用于描述自己是如何进行这项研究的
- Results 用于描述研究的现象和结构
- Conclusion/Discussion 用于升华主题，让具象问题变得抽象普适
而对于计算机会议论文来说，在章节方面会有一定程度的细化。Introduction 部分会包括：
- Background: 介绍背景知识
- Motivation: 介绍研究动机
Methodology 部分会包括：
- Insight: 完成设计的核心思想
- Design: 设计本身
- Implementation: 具体的实现
Results 部分会包括：
- 如果一个新技术，那么需要指明技术的应用场景 Use Case ；如果是原有技术改进，那么需要指明 Baseline
- Evaluation: 对设计的评估
Conclusion/Discussion 部分会包括：
- Discussion: 负责将研究升华，提供一些更加定性和抽象的描述
- Related Work: 用于和现有工作对比
- Conclusion: 对上文的总结回顾，起行文上的收束作用（而 Discussion 则承担设计上的发散升华作用）。
如果有机会，我希望按照更加符合计算机会议的方式重新组织一下这个模板。
Results
Revisit the Research Aim / Exisiting Research
不要直接描述自己的实验结果，这样过于唐突了，要想办法把读者和后面的实验联系起来，比较有效的方法就是回顾前文或者已有的其他人的工作，这些都是对实验的一种背景介绍，一种铺垫。
对前文的回顾：
as discussed previously, …
as mentioned earlier/before, ..
as outlined in the introduction, …
对实验目的的回顾：
in order to …, we examined …
it is important to reiterate that …
our aim/purpose/intention was to …
since/because …, we investigated …
to investigate …, we needed to …
the main purpose of this work was …
In this work, we sought to …
对现有其他实验的回顾：
as reported, …
it is known from the literature that …
the aforementioned theory/aim/prediction …
in earlier studies …
对结果的预期：
we reasoned/predicted that …
it was predicted that …
General Overview of Results
在介绍具体的实验之前，应当先给读者一个整体印象，这样才能避免读者陷入“流水帐式阅读”。整体印象应当包括我们的实验设置、实验目的、最终的实现效果。
对实验的总结概括：
in this section, we compare/evaluate/present …
the results are divided into two parts as follows:
对结果的总结概括：
generally speaking, …
in general, …
in most/all cases, …
in the main, …
it is appparent that in all/most/the majority of cases, …
it is evident from the results that …
on the whole, …
the overall response was …
using the method described above, we obtained …
Invitation to View Results
应当主动引导读者去看图或者看表，而不是只画图，而在正文里绝口不提。
对结果的被动引用：
(see Fig)
according to Fig
as detailed/evident/illustrated/indicated/listed/shown from/in Fig
as can be seen/found/identified/observed from/in Fig
comparing Fig.1 and Fig.4 shows that …
displayed in Fig
evidence for this is in Fig
inspection of Fig indicates …
be given/represented/visible in Fig
in Fig we compare/present ….
results are given in Fig
对结果的主动引用
Fig contains/corresponds/demostrates/displays/gives/illustates/lists/plots/presents/provides/reports/represents/reveals/shows/summaries
在引用结果时，某张图或者某张表的粒度可能过粗，可以更细一些，指向图中的某个具体数据：
the data in Fig
the rate in Table
the results of Fig
Special/Key Results in Detail
讲究的是客观描述与主观判断的结合。
Objective Description
这些词应当加上具体的数据来修饰，这样才能更有说服力和客观。比如说
2% increase
增加：
accelerate
expand
increase
be higher/highest
peak
precede
rise
减少：
decline
decrease
delay
drop
fall
be lower/lowest
reduce
变化：
change
vary
be different
produce
不变：
be constant/unaffected/unchanged
remain constant/the same
level off
存在：
exist
find
range from
be found/seen
occur
对比：
be equal/indentical/uniform
match
Subjective Description
数量评价：
abundant
adequate
inadequate
minimal
negligible
imperceptible
few
brief
sufficient
scarce
substantial
immense
程度评价：
almost
largely
likelihood
more or less
somewhat
most
mild
virtual
considerable
general
图表评价：
clear
measurable
sharp
smooth
steep
sudden
符合预期/不符合预期：
acceptable
appreciable
appropriate
unexpected
unusual
非常好/差：
dramatic
drastic
essential
excellent
extensive
important
marked
noticeable
overwhelming
poor
remarkable
satisfactory
serious
severe
significant
simple
striking
strong
suitable
surprising
radical
rapid
valuable
比较：
comparable
consistent
distinct
equivalent
excessive
exceptional
extreme
fair
obvious
only
resembling
similar
dominant
名词：
tendency
the majority of
main
Comparison with Other Results
和其他已有成果对比是非常必要的，baseline 必须好好选取，才能增加说服力。
印证/推翻：
as anticipated, …
as expected, …
as predicted by …
as reported by …
confirm
corroborate
disprove
inconsistent with
concur with
inline with
be in good agreement
prove
refute
reinforce
support
validate
verify
优劣：
compare well with
be better than
相似性：
consistent with
contrary to
match
be identical
be similar/dissimilar to
be parallel to
be unlike
correlate
Problems with Results
似乎这个会被写入 future work 而不是 evaluation 中。
问题：
immaterial
incompelte
infinitesimal
insignificant
less than ideal/perfect
a minor deficit/limitation
转折：
despite this
however
nevertheless
开脱：
at a preliminary attempt
neligible
not always/completely reliable/accurate/ideal/identical/clear/perfect/precise/significant
of no/little consequence/significance
only
reasonable results were obtain
room for improvement
slightly
a slight mismatch/limitation
somewhat
technicality
unimportant
原因：
may/could/might have been
beyond the scope of this study
be not the focus of this paper
be not within the scope of this paper
caused by
difficult to
due to
hard to control
inevitable
it should be noted that …
be not attempted/examined/expored in this study
possible sources of error
unavoidable
unexpected
unfortunately
unpredictable
unworkable
unavailable
解决办法：
futher work is planned
futher work should …
future work will
in future, care should be taken
in future, it is advised that
Possible Implications on Results
感觉似乎过于侧重 possible 了，有些让人感到虚弱了，不应该出现在这里。不应当成为一个必要的结构，而是应当成为一种行文的小修饰。
apparently
could be due to / explained by / attributed to / interpreted as / seen as
could account for
evidently
imply that
indicate that
in some circumstances
be owing to / associated with / likely / linked to / related to
it appears that
it could be concluded/infered/speculated/assumed that …
it is conceivable …
it is evident
it is logical that
it is thought/believed that
plausible
suggesting
likely
may/might/perhaps
possibly/possiblity
potentially
presumably
probably
provide compelling evidence
tend to / tendency